International Journal of Scientific Research in Engineering and Management (IJSREM)

Volume: 07 Issue: 05 | May - 2023 Impact Factor: 8.176 ISSN: 2582-3930

AI BASED ONLINE EXAM PROCTORING SYSTEM

Mr.S.Sankar Assistant Professor,Computer Science and Engineering , Dhirajlal Gandhi College of

Technology
Ms.J.Sherly Student, Computer Science and Engineering , Dhirajlal Gandhi College of Technology
Ms.K.Nandhini Student, Computer Science and Engineering , Dhirajlal Gandhi College of Technology
Ms.M.Swetha Student, Computer Science and Engineering , Dhirajlal Gandhi College of Technology
Ms.P.Sivaranjani Student, Computer Science and Engineering , Dhirajlal Gandhi College of Technology
---------------------------------------------------------------------***---------------------------------------------------------------------
I.ABSTRCT their methodology to authenticate the exam done from a
A weapon for enhancing one's life and helping nations thrive is distance. By minimizing human labour, our technology will
education. Moving towards e-learning and e-assessment is now affect online learning even after pandemics and can be a
necessary because to the extraordinary COVID-19 pandemic fantastic addition to the present online proctor suite.
breakout and country lock-downs. An engaging method of
testing distant pupils is through e-assessment. When deciding Key Words: E-assessment, Online exam proctoring, Temporal

whether to take examinations online, it is important to make convolution network, Electronic examination, templates,

sure that the test accurately assesses the knowledge, abilities, journals

and quantity of the candidate. Additionally, it is crucial to

II.INTRODUCTION
ensure the test is genuine legally, which means it is vital to stop
the student from getting outside advice or information. The Examination is a digital testing platform that enables students to
project's goal is to provide an online exam proctoring API that finish exams anytime they want. Exams are created in the
would allow students to take exams without being interrupted system in advance by teachers. The exam is conducted in a
or receiving information or recommendations from outside separate examination room with a computer and a camera
sources. The project's goal is to create an API for online exam monitoring system for recording. Additionally, there are
proctoring that would allow students to take tests without being sporadic inspections of the facilities. Exam exams are a good
distracted or obtaining unauthorized help. In this project, an substitute for the conventional classroom exams and are
end-to-end video-based deep learning method has been particularly appropriate for literature exams, maturity tests, and
proposed to identify the degree of at-risk students' participation personal exams. Exams increase study flexibility and relieve
in electronic written exams. This method feeds sequential video teachers of the burden of planning various examination
frames into a Temporal Convolution Network (TCN) for action occasions throughout the course of the academic year.
segmentation on videos to identify the degree of participation. Exams taken online provide flexibility for both teaching and
A online application called Online Exam Proctoring API, based learning. Additionally, they facilitate the use of a variety of
on the Model-View-Controller (MVC) architectural pattern, assessment techniques, facilitate the authoring and evaluation of
enables the management of electronic exams for specific answers, and lessen the amount of documentation required. The
courses. The system we created is broken down into many parts automated system eliminates the shortcomings of the
that analyse the video stream and ambient noise captured via conventional exam system and establishes the framework for the
the webcam, mouse, and keyboard actions. This system has new standard. Exams given electronically can be viewed as a
undergone a thorough re-engineering to accommodate the wonderful technological innovation. Exams that can be taken
difficult and complex task of conducting written exams online are considered to be electronic exams. These tests have
securely from a distance, including concerns with the appearance and action of a conventional exam. The sole
authentication, anti-cheating techniques, and a description of

© 2023, IJSREM | www.ijsrem.com | Page 1

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 07 Issue: 05 | May - 2023 Impact Factor: 8.176 ISSN: 2582-3930

distinction is that typing is required in place of handwriting for laborious physical labour, and are not ecologically friendly. To
the responses. A platform for electronic exams was created with verify applicants taking the exam, an invigilator is required to
the use of cutting-edge technology. This website was created be present at the exam Centre during the traditional exam
specifically to address the problems or shortfalls found in process. You require one invigilator for every 30 to 50
conventional classroom-based exams. candidates at each exam site. As a result, more than 25
invigilators would be required to supervise the exam of more
III. LITERATURE SURVEY
than 1000 applicants. This would entail excessive physical
In this research, a unique online proctoring system is proposed
labour and cost.
that employs deep learning to continuously monitor physical
locations without the need for a physical proctor to be there. B.PROPOSED SYSTEM
The system makes use of biometric techniques, including face
Develop an AI-monitored, remotely proctored exam with
recognition utilising the OpenCV and HOG face detectors. The
CamProctor utilising TCN is the project's suggested system.
technology also uses an eye blinking detection technique to find
Online proctoring systems frequently utilise Temporal
stationary images. Additionally, the technology is able to
Convolutional Network (TCN), a sort of deep learning
recognise devices like phones, laptops, iPads, and books to
technology, to identify suspicious activity during examinations
enforce fairness during tests. The FDDB (Face Detection Data
conducted online. TCN is a neural network architecture that
Set and Benchmark) and LFW (Labelled Faces in the Wild)
uses dilated convolutions to analyse sequential input, such as
datasets are used to test the system, which is implemented as a
time-series data. As a result, the online proctored tests track and
software system.Two elements make up the proposed web-
monitor students and aid institutions in eradicating exam
based online proctoring system. The online registration
cheating thanks to AI-powered technology.
component comes first, followed by Secondly, there is the
online proctoring component. Our online examination software provides auto proctoring
features, so your institute may conduct exams without fuss or
IV.SYSTEM IMPLEMENTATION concern about fraud or other irregularities. Online proctoring is

A.EXISTING SYSTEM an AI-based automated proctoring solution that automatically

invigilates exams using the webcams and microphones of the
Traditional tests are administered in classroom settings at
students.
designated testing locations. The following are some of the
difficulties involved with administering traditional exams. The B.1 Recognition of Faces
exam centres must be made accessible for the administration of
One of the crucial elements of the authentication process for
the exams. This includes expenses for locating testing
remote users throughout the online test process is face
locations, setting up the necessary infrastructure, and hiring
recognition.
examiners to oversee testing. To the testing locations, both the
candidates and the examiners must travel. Both time and money B.2 Face Recognition on CNN
may be wasted as a result of this. A sufficient number of the
Convolutional Neural Networks (CNN) are a widely utilised
paper-based answer sheets and question papers must be made
technique in many domains, including online proctoring
available. Exams administered on paper are therefore not very
systems, for face recognition and identification. Strong deep
environmentally friendly. Additionally, the answer sheets are
learning algorithms known as CNNs can automatically
physically transported to and from the exam sites and kept
identify and learn patterns from visual data.
somewhere. This method of travel is dangerous to misuse and
harm. Additionally, there would be expenses associated with B.3 Video Proctoring
infrastructure investment. Briefly said, traditional paper-and-
The footage is recorded as the webcam examines the students'
pencil exams are costly to administer and deliver, require
faces. If there is any unusual activity in the video, it raises red

© 2023, IJSREM | www.ijsrem.com | Page 2

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 07 Issue: 05 | May - 2023 Impact Factor: 8.176 ISSN: 2582-3930

flags. To assist you in determining whether an electronic test films. In an online exam, temporal convolutional neural
has been successfully finished with machine-driven remote networks were utilised to identify cheating. The education
invigilation proctoring, it would create an in-depth report for sector can be transformed by applying AI, and everything is
you. TCN enables AI-enabled remote proctoring systems to now virtual thanks to CamProctor. AI-integrated computer
continually learn, adjust, and advance in intelligence. systems can guarantee the validity of the exam by discouraging
cheating and the use of unfair tactics by the candidate. Using
B.4 Proctoring of images
Remote Proctoring, educational institutions don't have to
It evaluates and examines the photographs of the students, exams should be delayed or postponed during the disease
which are taken at regular intervals, to see if there has been any outbreak. Our suggested approach will help lessen inequality
cheating. Secure online exam environments are now possible throughout the online test. When using an online proctoring
thanks to image proctoring. system, human-induced detection is crucial since it helps
identify students' questionable activity throughout the exam.
B.5 Secure Browser
VII.REFRENCES
Prevents the candidate from opening multiple windows at once
while taking the electronic exam.Additionally, switching [1] X. Yang, D. Wu, X. Yi, J. H. M. Lee and Tan Lee, "iExam:
windows while taking the scheduled online exam is not allowed A Novel Online Exam Monitoring and Analysis System Based
with AI-powered remote proctoring. In order to maintain on Face Detection and Recognition", Electrical Engineering
control over the Online Exam environment, the protected and Systems Science, 2022.
browser technology required by the proctored examination
process applies this restriction. [2] Jiyou Jia and Yunfan He, "The design implementation and
pilot application of an intelligent online proctoring system for
V.ARCHITECTURE online exams", Interactive Technology and Smart Education,
2021.

[3] M. Ueda, T. Tanaka and H. Hasegawa, "Attempt to develop

a sound-based examination monitoring system" in Acoustical
Science and Technology, Acoustical Society of Japan, vol. 42,
no. 4, pp. 226-227, Jul. 2021.

[4] A.Nigam, R. Pasricha, T.Singh and P.Churi, "A Systematic

Review on AI-based Proctoring Systems: Past Present and
Future", Education and Information Technologies, vol. 26, no.
5, pp. 6421-6445, Sep. 2021.

[5] M.Labayen, R.Vea, J.Florez, N. Aginako and B. Sierra,

"Online Student Authentication and Proctoring System Based
VI.CONCLUSION
on Multimodal Biometrics Technology", IEEE Access, vol. 9,
Due to the widespread adoption of remote teaching and pp. 72398-72411, 2021.
learning, electronic tests are becoming a need. Recent exam
[6] K.Garg, K.Verma, K.Patidar, and N.Tejra, “Convolutional
proctoring solutions, however, are unable to offer a complete
neural network based virtual exam controller,” in 2020 4th
solution for preventing and identifying cheating during online
International Conference on Intelligent Computing and Control
exams. In this project, a Web dashboard called CamProctor was
Systems (ICICCS). IEEE, 2020, pp. 895–899.
created to identify exam-taking behaviour by using computer
vision techniques and deep learning algorithms to analyse exam

© 2023, IJSREM | www.ijsrem.com | Page 3

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 07 Issue: 05 | May - 2023 Impact Factor: 8.176 ISSN: 2582-3930

[7] Proctoredu, “Conduct credible online tests anywhere,

anytime.” https: //www.proctoredu.com/, accessed October 04, [17] R. Bawarith, D. Abdullah, D. Anas, and P. Dr., “E-exam
2020. Cheating Detection System,” International Journal of
Advanced Computer Science and Applications, vol. 8, no. 4,
[8] Comprobo, “Online automated invigilation or proctoring,”
pp. 176–181, 2017.
https://comprobo.co.uk, accessed October 04, 2020.

[9] M. Ghizlane, B. Hicham and F. H. Reda, "A New Model of [18] S. Prathish, A. N. S. and K. Bijlani, "An intelligent system
Automatic and Continuous Online Exam Monitoring", 2019 for online exam monitoring", 2016 International Conference on

Int. Conf. Syst. Collab. Big Data Internet Things & Secur., pp. Information Science (ICIS), pp. 138-143, 2016.

1-5, 2019.
[19] Yousef Atoum, Liping Chen, Alex X. Liu, D. H. Hsu
[10] H. S. G. Asep and Y. Bandung, "A Design of Continuous Stephen and Xiaoming Liu, Automated Online Exam
User Verification for Online Exam Proctoring on M-Learning", Proctoring, 2015.
Proc. Int. Conf. Electr. Eng. Informatics, vol. 2019-July, no.
July, pp. 284-289, 2019 [20] Ali Sharifara, Mohd Shafry Mohd Rahim and Yasaman
Anisi, A General Review of Human Face Detection Including
[11] O. Ruiz-Vivanco, A. Gonzalez-Eras, J. Cordero and L.
a Study of Neural Networks and Haar Feature-based Cascade
BarbaGuaman, "Monitoring for the evaluation process on-line
Classifier in Face Detection, 2014.
prototype based on openface algorithm", Communications in
Computer and Information Science, vol. 895, pp. 500-509,
[21] N.L Clarke, P. Dowland and S.M. Furnell, "e-Invigilator:
2019.
A biometric-based supervision system for e-Assessments",
International Conference on Information Society (iSociety
[12] G. Fenu, M. Marras, and L. Boratto, “A multi-biometric
2013), June 2013.
system for continuous student authentication in e-learning
platforms,” Pattern Recognition Letters, vol. 113, pp. 83–92,
2018.

[13] Z. Zhang, E. S. Aziz, S. Esche and C. Chassapis, "A

virtual proctor with biometric authentication for facilitating
distance education", Lecture Notes in Networks and Systems,
vol. 22, pp. 110-124, 2018.

[14] Y. Atoum, L. Chen, A. X. Liu, S. D. Hsu, and X. Liu,

“Automated online exam proctoring,” IEEE Transactions on
Multimedia, vol. 19, no. 7, pp. 1609–1624, 2017.

[15] Y. Atoum, L. Chen, A. X. Liu, S. D. H. Hsu and X. Liu,

"Automated Online Exam Proctoring", IEEE Trans.
Multimed., vol. 19, no. 7, pp. 1609-1624, 2017.

[16] R. Bawarith, D. Abdullah, D. Anas and P, "E-exam

Cheating Detection System", Int. J. Adv. Comput. Sci. Appl.,
vol. 8, no. 4, pp. 176-181, 2017.

© 2023, IJSREM | www.ijsrem.com | Page 4